EMA Day Trading Strategy: 9 EMA, 20 EMA/SMA, 200 EMA

Three lines do most of the work in EMA day trading: the 9, the 20, and the 200. The 9 EMA is your momentum line and your trigger. The 20 EMA (or 20 SMA, if you like it smoother) is the intraday trend you trade pullbacks into. The 200 EMA is the bias filter that keeps you on the right side of the bigger move. Nothing about those numbers is mathematically special. Part of why they matter is that so many traders watch them. StockCharts' ChartSchool guide to simple and exponential moving averages says the 200-day average may act as support or resistance because it is widely used, "almost like a self-fulfilling prophecy." That is a good reason to know the lines. It is not proof they work. Below is the setup, the math behind each line, how to test a setting before you trust it, and where the whole thing breaks.

What EMA Settings Should You Use for Day Trading?

An EMA is an exponential moving average, an average that weights recent candles more heavily than old ones. Investopedia's entry on the exponential moving average covers the definition. In practice it hugs price and turns faster than the equivalent simple moving average, which weights every candle in its lookback the same (the simple moving average guide has the SMA formula and a worked example). For day trading, that faster reaction is the point. When a stock gaps up and starts trending, the EMA tracks the move tighter than the SMA, so you get earlier pullback entries and earlier crossover signals. The cost is more whipsaws in chop, and most day traders accept that because they need the speed.

Here is how the three lines split the job. This table is the whole framework in one place, so it is worth reading top to bottom before the sections below.

How Is the EMA Calculated?

The math is short, and it explains why the 9 reacts to the last few candles while the 200 barely notices today. Each new EMA value blends the latest close with the previous EMA.

Schwab's thinkorswim reference for the ExpAverage function defines it the same way, with the smoothing coefficient equal to 2/(length + 1). For the 9 EMA, k = 2 / 10 = 0.20, so the newest close carries 20% of the weight. For the 20 EMA, k = 2 / 21 = 0.0952, about 9.5%, the same number the ChartSchool page works through. For the 200 EMA, k = 2 / 201 = 0.00995, about 1%.

A hypothetical candle makes it concrete. Say the 9, 20, and 200 EMA all sit at $50.00 and the next 5-minute candle closes at $51.00. The 9 moves to (51.00 × 0.20) + (50.00 × 0.80) = $50.20. The 20 moves to about $50.10. The 200 moves about a cent. Same candle, very different reactions.

The table runs the same arithmetic for every line here, plus the 21 and 50 from the alternates below. The last column is the useful one: if price jumps to a new level and stays there, how many candles until the line has covered half the gap? After n candles an EMA has covered 1 - (1 - k)^n of the jump. Plain arithmetic, not a backtest.

That is why the EMA feels faster in practice. The 9 gets halfway to a new price in about three candles. A 20 SMA, which just drops one old candle at a time, needs exactly ten. One quirk worth knowing: the very first EMA value has no previous EMA to blend with, and platforms handle that differently. StockCharts seeds it with a simple average. Schwab's thinkorswim doc starts from the first close. That, plus how much history is loaded and whether extended hours are included, is why two platforms can show slightly different values for the same 9 EMA.

If you want to see where these lines rank against everything else on the chart, the guide on the best technical indicators for day trading stacks moving averages up against VWAP, MACD, and RSI. And if you are running shorter timeframes, the scalp trading AI tool grades 9 EMA pullback setups on 1 to 5 minute charts.

Why the 9/20/200 EMA Combo, Not Other Periods?

The 9, 20, and 200 are not the only periods out there, and it is fair to ask why not an 8, a 21, or a 50 instead. Nothing in the math makes 9/20/200 better. What it has is reach. It is the set most day-trading education teaches (our indicator rankings included), so it is the set the most traders already have on their charts, and a level lots of people watch can attract more orders around it. That is a soft edge at best. A less common period can be just as valid and simply have fewer people looking at it.

None of these are "wrong." They are different crowds. The 8/21 pair (both Fibonacci numbers) is the alternate you will see most, often run as an 8/21/50 stack, and 9/21 is common too. If you already trade a different set and it works, the framework still applies: fast line for the trigger, mid line for the trend, slow line for the filter.

How to pick: start with the most-watched set. Change it only for a specific reason, like the fast line getting crossed every few candles on the market you trade, and test the change on replay before it touches real money. Then leave it alone for a few weeks. Swapping settings after every bad day is just fitting yesterday's chart.

What Is the 9 EMA and How Do You Trade It?

The 9 EMA is your momentum gauge. It averages the last 9 candles with the heaviest weight on the most recent ones (20% on the newest close, per the formula above). On a 5-minute chart that is roughly 45 minutes of price action. On a 1-minute chart it is 9 minutes. When price is running above the 9 EMA, buyers own the moment. When it drops below, sellers have taken short-term control. A stock grinding higher on above-average volume while holding above the 9 EMA on every pullback is showing you clean momentum. Some momentum traders call that riding the 9, and the momentum trading strategy guide covers how to find stocks moving like that in the first place.

The 9 EMA earns its keep in the first hour, when stocks move with conviction. During the lunchtime lull (a rule of thumb, roughly 11:30 a.m. to 1:00 p.m. ET) it chops back and forth and throws off false signals, so if a stock is drifting sideways through the 9 on thin volume, leave it alone. The 9 only means something when the stock is actually moving.

The most practical use is trailing your stop. When you are in a winning momentum trade and price is trending above the 9 EMA, move your stop to just under the 9 on each new candle. A close below the 9 is your signal that short-term momentum has flipped and it is time to take profit. That same idea, anchoring a stop to a line price actually respects instead of a round number, is the backbone of the methods in reading support and resistance.

9 EMA vs 20 SMA: Which Should You Use, and What Should You Set?

Neither average is better in general. The EMA turns sooner and throws more false signals, and the SMA is smoother and later. So you will see traders run a 9 EMA with a 20 EMA, and others run a 9 EMA with a 20 SMA. Both pairs work, and the difference is smaller than the internet arguments suggest. The 20 EMA reacts faster, so it gives tighter pullback entries and earlier signals. The 20 SMA is smoother, so it filters out some of the twitch and keeps you from getting faked out by a single ugly candle. If you tend to get chopped up jumping at every wiggle, the 20 SMA's extra smoothing is doing you a favor.

What to actually set: a 9 EMA on close, paired with either a 20 EMA or a 20 SMA on close, read on the 5-minute chart. Pick one of the 20s and stick with it. Flipping back and forth just adds decision fatigue. The trade rule does not change with the choice. The faster line (the 9) leads, the slower line (the 20) is your trend, and the cross between them is confirmation after price and volume already line up, never the trigger by itself. The same logic of comparing a price-reactive line against a slower reference shows up when you weigh VWAP vs EMA, where VWAP plays the institutional-anchor role the way the 20 plays the trend-anchor role here.

One thing to be honest about: in a steady trend the 20 EMA and 20 SMA sit close together, and the choice rarely decides a trade. They split apart after a sharp move, when the EMA covers half the gap in about 7 candles and the SMA needs 10 (the speed table above). That is three candles, 15 minutes on a 5-minute chart. If a setup only works on one of them and falls apart on the other, the setup was marginal to begin with. Spend the energy on volume and trend context, not on agonizing over which 20 to draw.

What Is the 9 and 20 EMA Strategy, Step by Step?

This is the setup people mean by the 9 and 20 EMA strategy, written as one ordered list so you can check it in sequence before you click buy. It is the long version. For a short, flip every condition: price below a falling 20, the 9 below the 20, the daily 200 overhead, a bounce into the 20 on light volume, and a trigger on a close back below the 9.

• 1. Trend
  Price is above a rising 20 EMA, the 9 is above the 20, and the daily 200 agrees (price above it for a long). If the 9 and 20 are flat and wound together, there is no trend to pull back into.
• 2. Pullback
  Price pulls back into the 20 on lighter volume than the push up. Heavy volume on the way down is real selling, not a pause. If price is stretched far above the 20 when you find it, there is no pullback yet, and buying it there is the chasing an overextended chart mistake.
• 3. Trigger
  Enter on a candle that closes back above the 9, or on a rejection bar off the 20 (a long lower wick with the close near the high). Wait for the close. An intrabar poke above the 9 that fades by the close is not a trigger.
• 4. Stop
  The initial stop goes below the pullback low. If that low sits right on the 20, put the stop a little under the line and treat the 20 as a zone, not an exact price. The guide to placing stops off chart structure covers how far under is far enough.
• 5. Partial
  Take partial size off at the prior swing high or at 1.5R to 2R, whichever comes first. Banking a piece here means a reversal costs you less of the move.
• 6. Trail
  Trail the rest under the 9 EMA, one candle at a time. A close below the 9 trims or ends the runner, and a close back below the 20 on volume ends it for sure.

The 20 EMA carries this setup. On a 5-minute chart it covers about 100 minutes, long enough to smooth out single-candle noise but short enough to track the day. The textbook version: a stock trends up through the morning, pulls back into the 20 on declining volume, then bounces with a volume spike. When that bounce lines up with VWAP at the same price, you have confluence, and confluence is what turns a B setup into an A. That same 20 EMA pullback often lines up with the VWAP momentum setups traders run intraday.

Test It Before You Trade It

Run the setup on replay or a paper account for a few sessions before real money touches it. Use the 9 and 20 exactly as written above, on the instrument and timeframe you will actually trade. Mark every pullback into the 20: which ones held, which ones broke, and what each trade would have returned in R from your entry and stop. That log tells you more than any settings debate, and it gives you your own win rate to hold up against the breakeven math in the win-rate section below. Replaying past sessions answers a different question from grading the chart in front of you right now, which the backtesting vs pre-trade grading breakdown walks through.

How Do You Trade 9/20 EMA Crossovers, and Where Do You Take Profit?

A crossover is just the faster EMA crossing the slower one. The 9 crossing above the 20 is bullish. The 9 crossing below the 20 is bearish. The 9/20 cross is the most-watched pair for day trading, and it works best as confirmation, not as a standalone entry. The cleanest version stacks three things at once: price pulls back to support, a reversal candle prints (a hammer, an engulfing bar), and the 9 EMA crosses back above the 20 right there. Three signals agreeing make a real case. The crossover alone, with no price-action context, fires far too often to trade.

It is sharpest in the first 90 minutes. After that, a lot of stocks chop sideways and the two EMAs crisscross without meaning. If you count three crossovers in 30 minutes, the stock is ranging, not trending, and the right move is to step away. You can stack a MACD read on top for extra confirmation, since MACD is built from the 12 and 26 EMA, so a MACD bullish cross at the same time as a 9/20 price cross is a stronger read than either alone.

Now the part people actually ask about: where do you take profit on a 9/20 crossover trade? The same way as the pullback. Take a partial at the prior swing high or 1.5R to 2R, then trail the rest under the 9 (steps 5 and 6 above). The crossover adds two exit rules of its own.

• Exit on a decisive close below the 9, then the 20
  A single close below the 9 EMA trims or exits the runner. A close back below the 20 EMA with volume means the intraday trend has changed character, so the trade is over even if you were still holding a sliver.
• Hard-stop the whole thing on the cross back
  If the 9 crosses back below the 20 (for a long), that is the inverse of your entry signal. No debate, you are out. The same line that got you in gets you out.

That exit structure, partial at a target then trail the rest behind a line price respects, is the same scale-and-trail logic that runs through the VWAP trading strategy, just anchored to the 9 EMA instead of VWAP.

What Is the Win Rate of the 9/20 EMA Strategy?

A bare cross traded through a chop zone loses far more than it wins. The same cross, filtered by the pullback-and-volume rules earlier in this guide and taken only in the stock's real trending hours, performs very differently. The periods are not the edge. The filter is, which is exactly why a single win-rate number for "the 9/20 EMA strategy" cannot exist in the first place, it would have to average across traders using wildly different filters.

Here is the number that is actually yours to calculate: the breakeven win rate for this guide's exit structure, a partial at roughly 1.8R average and the rest trailed behind the 9 EMA. The formula is 1 / (1 + R), where R is your average reward-to-risk across both exits. At 1.8R that is roughly 35.7%, so a 9/20 setup filtered this way only needs to work a little more than one time in three, a much lower bar than the high win-rate figures that circulate with no source attached. That is illustrative math for this exit structure, not a performance claim, since your real average R depends on where you scale out and how far you trail. The full derivation and a calculator live in the breakeven win rate section of the risk-reward guide. Hold your replay log up against that line instead of chasing a statistic that was never yours to begin with.

Is the 20 EMA / 200 EMA Still Good in Today's Markets?

Short version: yes, because neither line is trying to predict anything. The 20 EMA tracks the intraday trend, the 200 EMA marks the longer-term trend, and both jobs are just smoothed price. What people really mean when they ask if these still work is that faster, more algo-driven markets throw more whipsaws around any single line. True. The answer is the same as it has always been, do not trade the EMA naked.

The 20 EMA is the intraday trend anchor. Price above a rising 20 is a bullish intraday trend, price below a falling 20 is bearish, and the line acts as dynamic support in an uptrend and dynamic resistance in a downtrend. That is the job the step-by-step pullback above depends on, and nothing about modern markets has broken it.

The 200 EMA is a different animal: on the daily it represents roughly 10 months of price, and most day traders use it as a directional filter, not an entry. The rule is blunt. Above the daily 200 EMA, favor longs. Below it, favor shorts or sit out. A perfect 5-minute bull flag with the daily 200 EMA sitting overhead as resistance is a trade with a headwind. On the intraday chart the 200 EMA shows up less often, but when price does reach it, watch for a reaction. It tends to act like a magnet that either bounces price hard or gets sliced on heavy volume. Filtering your morning watchlist for stocks above their daily 200 EMA gives you a better pool to trade, because you are working with the bigger trend instead of against it.

Why the daily 200 and not a 50 on the intraday chart? Plenty of day traders run a 50 EMA on the 5-minute as a session trend filter (the speed table up top shows how much of the session it spans). That is a reasonable choice. This guide uses the daily 200 for bias because it answers a bigger question: is the stock in a longer-term uptrend at all? A 5-minute 50 can be rising nicely inside a stock that has been falling for months. If you want both, keep the daily 200 for direction and add the 5-minute 50 for whether today's session agrees.

The 50 and 200 on the daily are also the pair behind the two named crossovers: the 50-day average crossing above the 200-day is the golden cross, and crossing below is the death cross. For a day trader that is background context about the bigger trend, not an entry signal. The golden cross and death cross guide covers it.

What Timeframes Do the 9 and 20 EMA Work Best On?

The 5-minute chart is the home base for 9/20 EMA day trading. It balances signal quality against speed. The 1-minute makes the EMAs too noisy with constant false crossovers, useful only to fine-tune a fill once your 5-minute setup is already valid. The 15-minute is good for context but reacts too slowly for most intraday entries.

Can you trade the 9/20 on the 1-hour? Yes, but it stops being a day-trading tool and becomes a swing tool. On the 1-hour each EMA covers far more time, signals come slower, and you hold for hours or days instead of minutes. That is a different trade with a different risk profile, not a faster version of the intraday play. The clean workflow is multi-timeframe: read the daily and 1-hour for trend and bias, then drop to the 5-minute to time the entry against the same 9 and 20. Reading the same structure across timeframes is the core skill in the guide on how to read stock charts, just pointed at moving averages.

Does the 9/20 EMA Work on Futures, Forex and Crypto?

Yes. The formula does not care what it is averaging, so the 9, 20, and 200 calculate the same way on index futures, currency pairs, and coins as they do on stocks. What changes is the behavior around the lines. More volatile instruments cross the 9 more often, so the fast line whipsaws more, and some traders run a slightly longer fast line on those markets to cut the noise. If you change it, test the change the same way as above.

A few stock-specific parts of this guide do not carry over cleanly. Spot forex has no central exchange, so the volume on most forex charts is a tick count rather than traded size, which makes the volume checks in the pullback rules softer. Crypto trades around the clock, so "the first hour" depends on which session you treat as the open. Futures trade nearly 24 hours on weekdays, and whether your chart includes the overnight session changes where an intraday EMA sits. Pick which session your chart shows and keep it consistent.

If you grade these charts in SnapPChart, the read works the same way.

What Are the Most Common EMA Mistakes?

Most of the ways traders blow up an EMA strategy are not subtle. They are the same four every time, and they all come down to treating a line on the screen as a signal without the context that makes the line mean anything.

• Trading EMAs in a chop zone
  EMAs are trend-following. In a range, they get sliced on every other candle. If the 9 and 20 are tangled together and flat, the stock is not trending, and no crossover will save you. Wait for one side to win. Layering RSI on top can help you flag the sideways conditions before they cost you.
• Stacking too many EMAs
  The 5, 8, 9, 13, 20, 50, and 200 all on one chart is seven lines on top of price. Every candle touches something, so everything 'confirms.' Three lines (9, 20, 200), four if you add a 5-minute 50, is plenty. Past that you are adding noise, not signal.
• Ignoring volume at the EMA
  A pullback to the 20 EMA means nothing without volume. Heavy volume on the pullback is real selling and the level is likely to break. Volume drying up on the pullback and spiking on the bounce is the confirmation. The EMA is the level. Volume tells you whether anyone cares about it.
• Trading EMAs on dead stocks
  EMA setups want real participation. A rule of thumb (not a published threshold) is relative volume above roughly 2x plus real price movement. A name trading 100K shares a day in a 0.5% range will not respect the 9 EMA, because there are not enough participants to make the level real. Save EMA entries for stocks that are actually moving.

Combining the EMA read with a momentum oscillator like RSI is one of the cleaner ways to filter the chop-zone mistake before it costs you, and it is exactly the kind of multi-signal confluence that an AI chart analysis read pulls together in one pass.

Where Does AI Fit the EMA Read?

The honest version: AI does not invent a secret EMA. The 9, 20, and 200 are the same lines for everyone. What an upload removes is the two ways traders misread the setup under pressure. When you send a chart in, the read names which EMA or level price is actually reacting to, checks whether price is respecting the 9 and 20 or just poking through, reads the trend and market structure around it, and folds that into a single grade. So a thin-volume drift through the 9 gets flagged for what it is instead of getting chased because the candle looked exciting in the moment.

It is reading the screenshot, not your data feed. It judges what is on the chart you show it: the EMA positions, where price sits relative to them, the candle structure, the visible volume. It is not pulling order flow or recomputing the averages from raw ticks behind the scenes. That is the point, it grades the same picture you are looking at, just without the in-the-moment bias that makes you want every pullback to bounce. The fill and the follow-through are still the market's to give. The deeper version of how that grading works lives in the AI trading tools guide.